Gout affects over 5.1 million Americans, and recent studies have shown that the prevalence[unreadable] and incidence of gout in the United States (US) are increasing. Gout is characterized by recurrent attacks;[unreadable] these attacks cause pain, disability, and can eventually lead to joint damage. Recurrent gout attacks are[unreadable] often attributed to "triggers," precipitating risk factors that immediately precede an attack. Avoidance of[unreadable] these risk factors is a central preventive strategy for patients with gout. Nevertheless, much of the data[unreadable] identifying risk factors that may trigger recurrent gout attacks are either conjectural or irrelevant to[unreadable] contemporary populations. We propose to conduct an Internet-based case-crossover study to test a set of[unreadable] potential risk factors of gout with great relevance for contemporary adults.[unreadable] We will construct an interactive website for subject recruitment and data collection. We will recruit over 700[unreadable] individuals who have a diagnosis of gout, confirming their diagnosis by obtaining and reviewing their medical[unreadable] records. Using the Internet, we will systematically collect information on risk factors 48 hours preceding a[unreadable] gout attack (hazard period) and 48 hours preceding a day when no gout attack was experienced (control[unreadable] period). The study duration for each subject will be one year. Risk factors to be assessed include dairy[unreadable] product intake, quantity and type of alcohol consumption; acute infection and immunization, weather factors[unreadable] (i.e., temperature, barometric pressure, and humidity), and medication use (i.e., low-dose aspirin, statins,[unreadable] losartan, and allopurinol). A standardized questionnaire of gout symptoms and signs will be used to[unreadable] ascertain recurrent attacks that occur among participants during the study period. We will examine the[unreadable] relation of each factor to the risk of recurrent attack by comparing its frequency over the hazard period with[unreadable] that over the control period, and calculating odds ratio using the Mantel-Haenszel method. We will perform[unreadable] multiple conditional logistic regression models to examine the relation of each factor to the risk of recurrent[unreadable] attack while adjusting for other potential confounding factors. We will then assess whether the effect of a[unreadable] particular risk factor on the risk of recurrent gout attacks is modified by other risk factors.[unreadable] The findings from this study will have important implications for preventing recurrent gout flares as all risk[unreadable] factors we propose to study are common among patients with gout. If these recurrent flares were prevented,[unreadable] the burden of gout on society would be greatly alleviated. In addition, while focusing on specific risk factors[unreadable] for recurrent gout flares, the current proposal also represents the development of a methodological strategy[unreadable] to better understand why patients with recurrent rheumatic diseases experience disease flares.[unreadable]